A fast rank mutual information based decision tree and its implementation via Map‐Reduce

Abstract

Summary To address the time‐consuming problem for the confirmation of splitting attributes and splitting points in classic rank mutual information based decision trees, this paper establishes a fast rank mutual information based decision tree (FRMIDT) for classification problems. First, the proposed FRMIDT algorithm improves the velocity by a max‐relevance and min‐redundancy criterion to remove the redundant attributes in each tree node building. Then, the fuzzy c‐means algorithm is employed to confirm the splitting points for further acceleration. Meanwhile, a parallel implementation is developed in the framework of Map‐Reduce (MR‐FRMIDT) for medium or large‐scale data classification. Several comparative studies are conducted on UCI benchmark data sets. In contrast to the classic rank mutual information based decision tree on 12 data sets, the proposed FRMIDT model effectively reduces the computational time on the premise of keeping testing accuracy. Furthermore, the proposed FRMIDT algorithm is comparable through comparing FRMIDT with other traditional decision tree classifiers including BFT, C4.5, LAD, NBT, and SC. Meanwhile, the comparison with 7 different popular splitting measures based monotonic decision trees on several data sets illustrates the effectiveness of FRMIDT in monotonic classification. At last, the experimental analysis on other 6 data sets shows that the proposed MR‐FRMIDT is feasible and has a good parallel performance on reducing execution time and avoiding memory restrictions.

A parallel tree node splitting criterion for fuzzy decision trees

“…27,31 The cluster may contain thousands of computers and each computer offers a local storage and computation. 2,26 The greatest strength of MapReduce is decomposing the task on a big data into many small, single, and inherent tasks in many machine nodes of a cluster. 2,26 The greatest strength of MapReduce is decomposing the task on a big data into many small, single, and inherent tasks in many machine nodes of a cluster.…”

“…32 The MapReduce model supplies a brief programming framework for users to do large-scale data mining tasks in a distributed computing way, which can be implemented in many languages such as Java, C/C++, Ruby, and Python. 2,26 The greatest strength of MapReduce is decomposing the task on a big data into many small, single, and inherent tasks in many machine nodes of a cluster. Take the open-source software Hadoop 29 as an example.…”

“…2 Fuzzy decision trees have highly transparent and user-friendly characteristics 19 ; meanwhile, the amount of data produced in our daily life is showing a fleetness increasing tendency. However, when these methods are directly applied to deal with large-scale data sets, the competitiveness is lost due to some limiting factors such as memory capacity, execution time, and data complexity.…”

“…However, when these methods are directly applied to deal with large-scale data sets, the competitiveness is lost due to some limiting factors such as memory capacity, execution time, and data complexity. 2 Fuzzy decision trees have highly transparent and user-friendly characteristics 19 ; meanwhile, the amount of data produced in our daily life is showing a fleetness increasing tendency. [20][21][22][23][24][25] Therefore, it is unavoidable to learn a classifier based on fuzzy decision trees for dealing with large-scale data sets.…”

An ExtensionDecision Tree Algorithm for Lightweight Design of Autobody Structure

“…The decision tree (DT) is a popular classification and prediction algorithm [15][16][17][18][19]. The DT can extract the relationship between data features and classes in a set of disordered, irregular cases, and, on this basis, classify the current data in an accurate manner [20][21][22][23].…”